# Appendix III: Hmmmmmmmmm¶

The program developed has been run on two different machines. Two plots of runtime versus data size with various numbers of processors are presented.

## LittleFE¶

The LittleFE is a system with $$6$$ dual core CPUs connected by a 1Mb ethernet. Each CPU is housed on its own board. Each board is connected to a 1Mb ethernet. Each of the CPUs is an Intel Atom D525 with XXGB of ram.

## Questions¶

1. Carefully examine this plot.
2. Approximate the speedups when using $$2$$, $$4$$, $$8$$, or $$12$$ processors.
3. Does this fit into the overall analysis presented earlier in this module?
4. What can you now tell me about some hidden constants? Are they significantly important?
5. Suppose your parallel processors are interconnected by a switched architecture. What would you expect?

## Dell PowerEdge C6145¶

The Dell PowerEdge C6145 system is a with $$4$$ $$16$$ core CPUs on one board. Each of the CPUs is an AMD Opteron 6200 with XXGB of ram.

## Questions¶

1. Carefully examine this plot.
2. Approximate the speedups when using $$2$$, $$4$$, $$8$$, $$16$$, or $$32$$ processors.
3. Does this fit into the overall analysis presented earlier in this module?
4. What can you now tell me about some hidden constants? Are they significantly important?

#### Previous topic

Appendix II: Utilities